Auto-scale cloud-based memory-intensive applications by sizing a machine's physical memory correctly for critical applications

Cloud providers offer simplistic scaling policies that rely on thresholds which force tenants to have a prior knowledge of their workloads causing them to undersize or oversize the optimum needed memory. Even a small decrease in the amount of memory available to an application can have a dramatic impact on performance, much more than CPU power.

Researchers have develop a new method for scaling memory-intensive workloads by finding an optimum operating point without setting a specific threshold. This makes it worry-free for tenants, and continually adapts and fine-tunes allocations as workloads evolves, optimizing cloud storage resources and increasing overall processing speeds.

Operational Demonstration

• Determines a natural threshold for memory-intensive applications

• Effective in horizontally and vertically scaling memory intensive workloads.

• Save on operating costs while avoiding queuing, thrashing, or collapse.

• Improve overall processing speed.

J. Novak, Sneha K. Kasera (2017), Auto-tuning Active Queue Management, in Proc. of 9th International Conference on Communication Systems and Networks (COMSNETS), January 2017.

https://ieeexplore.ieee.org/document/7945369

Publication Number: US-2020-0218574-A1
Patent Title: AUTO-SCALING CLOUD-BASED MEMORY-INTENSIVE APPLICATIONS
Jurisdiction/Country: United States
Application Type: Non-Provisional

Jonathan Tyler
801-587-0515
jonathan.tyler@utah.edu